wanted to know about building
stuff but were afraid to ask."
Marshall G. Emm N1FN/VK5FN
[This series was
originally published in "73 Amateur
Radio" between November 1997 and February 1998]
(You are Here!)
Basic Tools Table
Populating the Board
The Smoke Test
Resistor Color Code
Tools You'll Need
Preparing the Enclosure
Creating "Panel Art"
Isolate the Problem
Find and Fix
A Horror Story
This is the last article in the series on basic electronic construction, and I hope you don't even need to read it! If you been with me from the start, you probably have a brand new, working, VM-110 AC Voltage Monitor(1), installed neatly in an enclosure with a nice label.
But what if it doesn't work? Don't despair. The odds are very good that you can figure out what is wrong and fix it yourself, through a process we mistake-makers refer to as "trouble-shooting."
Exhaustive research(2) has shown that out of every 1000 kits which fail to work, exactly 985 times the problem is a mistake by the builder. Exactly 7 times it is a minor manufacturing defect in a printed circuit board. 5 times in a thousand it's a faulty component. And exactly 3 times out of a thousand it is a mistake by the designer of the kit (most often when they have made a change "between versions" and missed some minor detail, and that kind of problem only lasts for as long as it takes one person to build the kit and send it back).
So a sensible approach is to ask yourself, "what did I do wrong." If you can determine that you did nothing wrong, look for a problem in the circuit board, and then a faulty component. If you still haven't turned up the problem it's time to send it back to its maker. Odds are pretty good you'll get it back with a polite note that it was your fault after all-- they found a goof that you missed in the umpty-teen times you checked everything.
Step One-- Take a Break!
Seriously! You are probably "too close" to your project and if you saw something one way when you installed it you will probably see it the same way when you check it, unless you have let a bit of time go by and can approach the problem with a fresh eye.
People who write for a living often think they can adequately proof-read their own material. People who edit for a living know better. And I know from personal experience that I often find mistakes in an article when in comes back in proof form. It takes awhile for that to happen, so my eyes see what is actually on the page rather than what I "think" I put there.
Exactly the same principle applies in electronic construction. You checked the diode when you installed it, and it looked right when you checked it. Forty-eight hours later you're wondering how you could possibly have put the darned thing in backwards.
There is also a certain amount of anxiety associated with trouble-shooting your own work. You put a lot of time and effort into building it, not to mention the dollar cost, and it's all wasted if the thing doesn't work. This anxiety makes it easy to jump to conclusions, take short-cuts in checking your work, and seek someone else to put the blame on.
If you just heave a big sigh and put your project away for a day or two before trying to fix it, you will come back to it with a fresh eye and a good attitude, and much better prospects for success.
Let me give you a concrete example. I recently finished a project late one night and was just devastated when it didn't work when I flipped the switch. My immediate inclination was to tear into it and fix it immediately, but instead I heaved a big sigh, put it on the shelf, and forgot about for two or three days. When I came back to it... well, it, um... it worked. I'd only made one tiny mistake the other night-- I forgot to actually turn the power supply on. OK, my face is a little red over that but maybe it will help you to avoid similar problems.
The procedures I am going to describe are just guidelines. It's a system that works for me, but I take liberties with it sometimes and you shouldn't feel that you have to do everything I'm going to describe, or do it in the same order.
Don't be embarrassed to ask for help. If you know somebody who is a builder, ask him to cast an eye over your work. You might be surprised how easily someone else can see a problem that you have overlooked. And most of your peers are going to be delighted to be asked, because most hams are helpful and it's also an opportunity to show off.
Lazy Boy Step Two-- check all the components and connections. I call this the Lazy Boy approach because it is easy, but it is also time consuming. You check all of the components and soldering (but don't try to do both at the same time).
In checking the components, you need to check three things-- the value, the orientation (for polarized components like diodes and electrolytic caps), and that they are installed in the correct location. Go through the instructions step by step and check each step off as you go through the sequence. You may also wish to consider going through the instructions in reverse order (just like adding a column of numbers), to give you a slightly different perspective.
|Rule #1: Murphy was right.
If you check
everything, the problem is bound to be the
last thing that you check.
In checking the soldering, use as strong a magnifying glass as you can find (a 10X loupe is perfect) under good light. What you are looking for are poor connections (cold joints, or connections with a rough finish or no sign of solder flowing up the component lead) and solder bridges-- solder which has flowed between adjacent tracks or connection points. Refer to the solder track diagram if one is available when you can't be sure whether a connection is intended. If you don't have the solder track artwork you will have to refer back to the schematic, which can be a real pain. So here's a hint for your NEXT project-- if a copy of the art is not supplied, photocopy the solder side of the circuit board before you start. When you are checking the soldering, you are looking at mechanical detail without any regard for "where you are" in the circuit-- start at one corner of the board and work your way through to the other side in reasonable stages so you can be sure you have examined every square millimeter.
Sometimes you will encounter a "whisker" solder bridge or the tiniest trace of circuit board material across two tracks. Use your hobby knife, or the edge of a small screwdriver, or (best of all if you have one) a dental pick to scrape the board material between the tracks to remove the bridge. Often I have been able literally feel a bit of solder or conductive material that was all but invisible to the naked eye.
If you have checked all the components and all of the soldering and it you haven't found anything to fix, you have two choices-- send it back to the maker or start over with the method outlined below.
Probably you will get your circuit working
with the above approach,
but Murphy's law says no matter what order you check your work in, the
problem will turn out to be in the very last component or joint that
check. You can save a lot of time by applying the following method,
especially with more complex circuits.
The REAL Step Two-- Isolate the Problem
You don't have to be an electronic engineer to have some idea of what the various parts of the circuit do. The instructions probably tell you in general terms, and common sense can be pretty useful to. And it's quite likely that some parts of the circuit are working perfectly. We're going to go through this in stages, and in some cases if you detect a problem you will immediately see the cause and the solution. But if not, don't despair-- isolating the problem is only the first of a lot of things we can do.
|Safety First! Working
circuitry can be dangerous. Know what you
are doing, and err on the side of caution.
|Rule #2: Follow the POWER, and follow the SIGNAL!|
To check power supply and usage characteristics, it helps to have a chart of voltages at various test points such as the pins of integrated circuit chips. If you project included such a chart, you probably won't need much more. Just check the indicated voltages and look for a problem in the immediate area of the test point if the voltage is not as specified.
If you don't have a chart of reference voltages, you can still achieve much the same effect by tracing the power supply lines on the schematic diagram. It's quite possible that you won't know exactly what voltage should appear at a given point, but if there is no voltage at all you have located a problem.
You can also check for continuity between many integrated circuit pins and ground (with the power disconnected, of course). Look at the circuit diagram and find the grounding points and check them. An open or missing connection to ground can be just as problematic as an "extra" one!
To follow the signal, you probably need to know a bit more about how the circuit works. There are three kinds of signals that you may be able work with-- audio, RF, and digital.
Audio signal tracing can be done with a pair of headphones connected to a pair of test leads. Probe the circuit at various points along the signal path by connecting the grounded or :"common" side of the headphones to the circuit board ground, and the "hot" connection (usually the tip on a standard phone plug) to the test point. You can start at either end of the audio path. If you are starting at the source, you follow the signal until it disappears, which tells you approximately where the problem is. If you are starting at the output, you follow the path until the signal appears, which tells you the same thing.
RF signal tracing is a bit more difficult. Two possible approaches are to use a simple diode and capacitor circuit as detailed in ARRL Handbook (or of course an RF voltage probe if you have one). The probe circuit rectifies the RF and produces a DC voltage which can be measured with your multimeter. Be careful, though, not to confuse RF and DC voltages, because both will be present at some points in the circuit.
The second approach is to bypass suspect stages. If you are working with a receiver, for example, you can find the input and output points for a filter, and use a clip lead to bypass the entire filter. If you suddenly have a working receiver, you know the problem is in the filter stage.
Digital signal tracing is generally beyond the scope of this
but in many simple circuits you can easily determine whether a specific
point should be "on or high" (commonly +5V) or "off or low" (commonly
than +1V). For example, the keying output of an electronic keyer is
by one specific pin on one specific chip, which is "high" when a dit or
dah is being sent.
Step Three -- Clean it Up!
No matter how carefully you solder, there is bound to be some flux residue on the track side of the board. Depending on how much there is, and what type of flux it is, it may make it difficult to see problems. I know I said (in part one, November 1997) you probably shouldn't bother removing the flux, but a clean board is a lot easier to work with when you are trying to find a problem. It's fine with me if you want to clean the board before performing step two (above), but keep in mind that cleaning the board may introduce problems that weren't there before.
If you are using ordinary resin-core solder, you can buy a very expensive solder flux remover at your local radio parts store, but the easiest material to remove excess flux with is acetone. Acetone is probably better known (and certainly more readily available) as nail polish remover. You'll probably find several varieties in your local grocery or drug store, and this is one of those cases where a brand name means nothing. Buy the cheapest they have, but do look for one that is described as "non-oily" or lists only acetone on the ingredients label. This time we're lucky, because it's usually the more expensive "name brands" that have the added oils, "emoluents," and perfumes. If you can't find a pure acetone nail polish remover on the shelf, ask the pharmacist-- he'll probably sell you a small bottle of it at a very reasonable price. And don't overlook your neighborhood hardware super-store, where you'll probably find a quart bottle of acetone for about the same price as four ounces of nail polish remover.
|Warning! Acetone is
and you do not want to breathe the
vapors. Use only in an open area away
from any flames, and please don’t smoke
while you’re working with it. Wear rubber
If you are not using resin core solder you should use whatever solvent is recommended by the manufacturer. If you are using a water soluble flux, for example, you can clean the board with warm water (with a small amount of detergent in it) but you must expect it will take longer for the board to dry than if you use a volatile solvent. And you must wait until the board is thoroughly dry before trying to do anything else with it.
Step Four-- Find and Fix.
Assuming you have localized the problem, or determined that it is in a particular area of the circuit, it should be relatively easy to narrow it down even further to a specific component or circuit path.
Since the odds are still pretty good that the problem is a soldering fault, you might want to try the "wiggle test." Basically you just press on each component with a fingertip and wiggle it a bit to see if the circuit suddenly starts to function. Note this should only be done with low power circuits! In fact, you should probably not use your fingertip in the first place. Use the eraser end of a pencil, or a non-conducting tuning tool.
Look at the components and the soldering again, and see if you can find:
Substitute another component. If you happen to have another component of the same value, you can swap it in and see if it fixes the problem. Often you can use a "close" value as a temporary substitute if you can't find an exact match, and if you are lucky that will tell you whether the original component was in fact faulty.
Bridge the component. You need to be careful with this one, especially with components in the power circuit, but often you can safely bypass a component long enough to see if doing so will make the circuit function, in which case the component is probably faulty.
Lazy Boy Step Five.
If you are tired of messing with it, and have the money, and/or are just downright lazy, by all means pack it up and send it back to the manufacturer. Check the documentation for instructions which will often tell you the specific fee that must be sent, or whether you need to arrange for a "return authorization."
The Real Step Five.
Don't give up just yet! If you've got this far and your project still doesn't work, it is time to get in touch with the manufacturer, but you may still be able to fix it with a bit of help. Call them, or send them an e-mail, or write to them, with as complete a description of the problem as possible, and what you have done to try to fix it.
Most people who sell kits to (or design projects for) the amateur radio community do so as a labor of love, and they want you to succeed! Many of them will fix a non-functioning kit without charge, or at a very reasonable cost, and often they will send you replacement parts or spend so much time trying to help you that they couldn't possibly break even on that particular sale. Here are some of the reasons you should get in touch before sending it back:
Most suppliers will meet you half-way, too. If the problem was their fault, they won't charge you for the repair (and some won't even charge you the freight to get it back to you). If it was your fault, then you certainly shouldn't mind paying a reasonable price for a remedy. In all my years of building kits, I have had:
one that I didn't bother sending back because it was cheaper and easier to order another one and start over (the second one worked and gave me a comparison unit to get the first one going!), and
only one that resulted in an irreconcilable difference of opinion and a demand for a refund.
That one unfortunate situation is worth dwelling on as an example of how bad things can get, but please note that these were folks whose kits have nothing to do with ham radio and therefore will probably not be encountered by most of you. I won't name them, because I'm sure they have either cleaned up their act by now or perhaps gone out of business. The kit was a power supply, and the problem was a faulty pass transistor AND probably a fault in the basic design. Using most of the techniques described above, I determined that I had built it properly, but the malfunction pointed to the transistor. I swapped another one in and it still didn't work, so I decided to send it back after speaking to one of their "technical" people. There wasn't a whole lot of communication because the guy didn't speak much English, but he did authorize me to send it back on the understanding that I would pay if the problem was my fault. There was just something the slightest bit "off" about these people, so I put the original transistor back in, double checked the soldering, and photographed both sides of the board before sending it back.
They "fixed" it, charged my credit card $35 (the kit cost $29 to start with!), and said the problem was a cold solder joint on the pass transistor. Well, guess what?
I had a close-up photograph of the other side of the board that showed they had in fact replaced the pass transistor.
And, to put the icing on the cake, the &*#^ thing still didn't work.
The moral of the story? There are at least two:
As we've worked our way through this four part series, your kit probably worked first time. If it didn't, you were probably able to fix it all by yourself. If you couldn't, then you've probably sent it back and got it fixed, perhaps paying a fee which you can regard as a tuition payment. In any case, you're finished with it now and can start thinking about that next project-- perhaps a keyer, or an audio amplifier, or even a transceiver!
As I write this I have already started to receive a considerable amount of correspondence resulting from part one back in the November issue of "73." I understand many of you have even sent in the "feedback card" (thanks!). When I started this series I was concerned that it would be difficult to find the right level of detail for the beginning kit builder, but I am highly gratified by the response and just thrilled that it has convinced some of you that you can do it!
I hope when you build your next project you will write it up for publication. You don't have to be an expert, and you don't have to be a "writer." What counts is your ability and willingness to share your experience. HOPE CU SOON IN PRINT ES 73 DE N1FN.
1. Rainbow Kit VM-110, $10.95 (+$5 s/h) available from Milestone Technologies, 10691 E. Bethany Dr., Suite 800, Aurora, CO 80014-2670 (303) 752-3382, orders (800)238-8205 email: email@example.com, http://www.mtechnologies.com. Also available from Electronic Rainbow Inc., 6227 Coffman Rd., Indianapolis, IN 46268 or call 317-291-7262.
2. Conducted by the author, who was mainly exhausted by the effort involved in getting the numbers to add up to 100%